TAUVEX and UV Transients
From Tauwiki
== The TAUVEX: Mission Overview and Prospects for Observations of Transient UV Events ==
\author{Jayant Murthy\inst{1} \and Margarita Safonova\inst{1} \and C.~Sivaram \inst{1}} \institute{Indian Institute of Astrophysics, Koramangala 4th block, Bangalore 560012, India}
Contents |
[edit] Abstract:
The TAUVEX (Tel Aviv University Ultraviolet Explorer) is a UV imaging experiment that will image large parts of the sky in the wavelength region between 1400 and 3200 \AA. \T is a collaborative effort between the Indian Institute of Astrophysics and Tel Aviv University, and is scheduled for early 2008 launch and at least three years of operations. The scientific instrument has been fabricated at ElOp in Israel with the satellite interfaces and the launch provided by ISRO. ISRO will also perform the flight operations. India is contributing to the pipeline development and main mission planning is being conducted at the Indian Institute of Astrophysics. The \T Science Team have created a coherent observing program to address several key science objectives. Much of the \T time will be dedicated to these observations. However, \T Science Team will also pursue a number of moderate-size projects designed to study specific astronomical objects or phenomena, one of them is the short-scale UV transient events. In this paper, we outline some of the TST plans for observing the UV flares. We also present a description of the \T mission, including instrument design and its estimated performance.
[edit] Introduction
With the launch of the \T (Tel Aviv University UV Explorer) early next, 2008, year, the astronomical community will have access to the most advanced attempt to operate a flexible instrument for observations in the mid-UV range $135-400$ nm. \T will have a point source sensitivity much better than GALEX (by about $5^m$) in $20\%$ of the sky and in three UV bands compared to two of GALEX. With a $0.9^{\circ}$ field of view (FOV), \T is well-suited for observing nearby extended objects, impossible for example, for {\it HST} with its greater resolution but narrow field of view. The three parallel telescopes of \T will enable to image simultaneously at two or three different frequency bands, thus measuring the UV colours of all or most objects in the image frame. However, \T sources will also include nebulae of various types, galaxies and clusters of stars and galaxies.
Frames obtained by \T will also contain valuable information about UV background in the Solar System and in the Galaxy. The UV sky background is correlated with the Galactic HI column density, and the lowest acceptable value is derived to be about 100 ph/sec/cm$^2$/\AA/steradian \citep{noah-IAU168-94}. It is possible that \T will be able to detect the lowest UV background value quoted above with a S/N$\ge 5$. Using three different filters, it may be possible to obtain the wavelength distribution of the UV sky background. The main limitation of \T will be a straylight, especially solar straylight at angles less than $90^{\circ}$ with the line of sight (LOS). In parts of the orbit with no straylight, however, the detection will be limited by only the photon statistics, as the \T detectors are virtually noiseless \citep{{noah-phys-scripta98},{TOM}}. Effective exposure time of about 1000 sec per field in each of three \T bands will yield the resultant monochromatic UV magnitude of 25, or almost 15 magnitudes better than the $TD1$ survey and with much better spatial resolution.
\T is well-suited for UV variability studies as well, the photometry capabilities (each detected photon is time-tagged with a precision of $\sim 128$ ms) and the possibility of repeated scans will enable the recording of UV light curves of variable sources; however, the time scale at which variability can be investigated would depend on the source latitude and brightness. Here also \T has an advantage over GALEX, as GALEX does not specifically study variability, though serendipitous discoveries are possible in both instruments, GALEX already recording several short-time scale flares \citep{galex-flares}.
[edit] \T Team Science Investigations
The main scientific goal will be to survey the sky over the mission lifetime combined with selected deep fields and observations of interesting targets. The limiting magnitude will be on the order of $19^m$ for the UV 3-band survey and $25^m$ for the deep pointings. (Magnitudes quated are as per Hayes \& Latham \cite{monochromatic}.) \T will certainly detect more than $10^7$ galaxies with three-band photometry plus a comparable number of stars and, perhaps most excitingly, several million quasars. The deep census of a large fraction of the sky will yield numerous newly discovered objects which will be followed up from ground-based observatories in India and Israel for identification and further classification.
To this end, the TST has created several key science areas some of which are outlined below.
[edit] Deep Sky Surveys
Nominal mode of \T operation is scanning the sky along a fixed declination with complete sky coverage achieved by rotating the platform, on which \T is mounted, from $-90^{\circ}$ to $+90^{\circ}$. About 80\% of observing time was chosen to be an all-sky survey. GALEX will observe a large fraction of the sky, or perhaps even the entire sky, as part of its sky survey and will go deeper in selected areas such as the Hubble Deep Field. We have chosen as our operational policy during the first 18 months to observe some fraction of the sky---perhaps 15\%.---to a depth much greater than GALEX. In order to achieve more significant results than those of the GALEX MIS (our reference comparison) we must cover a solid angle 2.5 times larger, thus survey 3250 square degrees. This is the reason that the first-year proposed survey goal is to cover with deep exposures, of $\sim 5000$ sec per source, approximately 1500 square degrees around each Celestial Pole (where the exposure time theoretically reaches 86,000 sec a day). The important difference in comparison with GALEX is that this survey would be done with three UV spectral bands instead of two.
\subsection{Diffuse UV Radiation}
The observed diffuse UV radiation field is almost entirely due to starlight scattered by interstellar dust. With its wide field of view ($0.9^{\circ}$) combined with the ability to reject stars because of its 2-dimensional imaging capability, TAUVEX will be able to trace the level of the astronomical diffuse radiation over the sky. The dark count rate from the instrument (the rate when no astronomical source is being observed) is very small and will be measured at intervals using a blocking filter. Terrestrial emission which often affects satellites in lower orbits will simply not be present at geostationary altitudes (except for the Lyman lines of hydrogen, which will be blocked by the filters). Particularly in the near UV, zodiacal light---light from the Sun scattered by dust within our Solar System---will compete with the diffuse Galactic light and will have to be modeled and subtracted. Because \T will be observing stars over wide swaths of a Galaxy, it will be possible to map the interstellar extinction over much of the sky.
[edit] Galaxy Surveys
Nearby galaxies are so large that they cannot be observed in their entirety by an instrument such as the HST with its exquisite resolution over a small field of view. Thus we know relatively little about the star formation rate at low redshift---a topic which will be addressed by TAUVEX.
In a survey mode TAUVEX will use primarily its three principal filters SF1, SF2 and SF3, which span the spectral region from somewhat longer than Lyman $\a$ to 320 nm with three well-defined bands. Three filters define two colour indices in the UV and the combination of these measurements, with data from the optical and infrared, allows the derivation of even more colour indices. As known from the optical domain, colour indices are magnitude differences, representing spectral slopes, and these are indicative of the mechanisms producing the radiation.
Since the stellar population in a galaxy evolves, the overall shape of the spectral energy distribution (SED) changes with time. The peak of the SED of a single stellar population becomes redder as time passes since the last major starburst event, moving from the far-UV to the near-UV then to the optical. The use of three UV bands that sample the SED peaks of B, A, and F-type stars will allow, in conjunction with optical information (e.g., from a number of SDSS bands), proper accounting for the stellar populations composing the galaxies covered by the survey. It would be possible to disentangle star formation histories during the last few Gyrs and reconstruct the influence of the neighborhood on the evolution of galaxies in clusters, groups, and in the field.
Other very interesting TAUVEX targets are the two Magellanic Clouds and the stellar and gas bridge linking them (the Magellanic Bridge). The LMC and SMC were observed already with the rocket-borne telescope flown by Smith et al. \cite{smith}, with the FAUST telescope on the Space Shuttle \citep{FAUST}, and parts were observed with UIT \citep{UIT-parker} and HST \citep{elhanan2005}. While the HST observations have an exquisite angular resolution and resolve individual stars, the two first observations sample arcmin-sized regions and only the UIT images offer an angular resolution somewhat similar, but inferior, to what TAUVEX will produce.
[edit] Variability Studies
Because TAUVEX will continually scan over the sky, it is well suited for detecting (measuring) variability on various time scales ($<1$ day--3 years). Advantages for TAUVEX are that there is a large number of sources in these fields with full spectroscopic follow-up and that there is no UV and variability information currently available on them. For example, the AGN phenomenon manifests itself very strongly in the UV, and TAUVEX will be used to determine their spatial, broadband spectral and temporal structure (i.e., variability on scales hours to months).
[edit] Focused Investigations
The first science operations immediately after the IOC phase will be the First Science Survey (FSS), which is intended to provide an early examination of the UV sky. The primary purpose is to provide an early and representative sample of reliable UV data that will enable effective and efficient planning of \T observing programs. The survey will characterize both high- and low-latitude regions of the sky, and will access the effects of zodiacal emission. The FSS will include both Galactic and extragalactic components. The observational planning for the FSS will be available on the \T web pages.
In addition to the main key science objectives, \T Science Team will conduct studies in a number of areas using data from the key programs as well as supplemental observations. These investigations will include searchers for UV flashes from planetary collisions, UV observations of selected ULXs, searchers for the UV emission from pulsars and so on.
[edit] \T and UV Emission from Celestial Collisions
The phenomena of one celestial body smashing into anther is quite common on all astronomical scales, from asteroids bombarding planets to recently discovered collisions of massive galaxy clusters \cite{{bullet}, {cl0024}}. The origin of blue stragglers in the globular clusters is due to merging of two, or, may be three, main sequence (MS) stars of $\sim 0.8$ \msun. Perhaps, about half the stars in central region of some globulars underwent one or two collisions over a period of $10^{10}$ years \citep{sivaram-ref3}. White dwarf (WD) binaries with periods $< 5$ mins will merge in a few thousand years, and neutron star--white dwarf (NS--WD) binaries with periods of $\sim 10$ mins have been observed \citep{sivaram-ref4}.
Over the past decades, a standard model for planetary system formation has emerged, in which the final stage is now strongly associated with a significant number of giant impacts on each of the young forming planets \cite{stern}. Giant Impact Models (GIM) are often invoked for explaining the formation of planetary systems, the Mars-sized terrestrial impactor believed to be responsible for the formation of the Moon was such an impact \cite{stevenson87}. However, even at later stages of planetary system life, the giant collisions are possible. In the past decade about 200 extra-solar planets (ESP) have been detected and they are a rapidly growing sample. Most of ESPs are giant hot jupiters, locked in a very short-period orbits with their star. It is reasonable to expect planetary and stellar-planetary collisions in such systems.
[edit] Electromagnetic Transient Events in UV
Electromagnetic Transient Events include X-ray bursts, flashes, X-ray flares, GRBs, etc. The discovery and study of highly transient sources, especially those that rise to high brightness and then fade, has been a major part of modern astrophysics. A lot of studies look at them from either one extreme end of the electromagnetic spectra, as very high-energy phenomena, or from the other one, infrared (IR). For example, even before the discovery of ESPs around MS stars, Stern \cite{stern} was discussing the possibility of detection of planets through IR signals from their collisions. However, the UV part of the spectrum was largerly ignored. Recently, it was proposed that planet-planet collisions will give rise to flares, where at the flash peak time, the EUV flash can greately outshine the host star \citep{zhang}. On the observational side, 2006 and 2007 years have been witnesses to the detection of UV flares from M-dwarfs by Deep Lens Survey (DLS) \citep{DLS} and GALEX \citep{galex-flares} and to a serendipitious discovery of an energetic UV flare from a UV Ceti star by GALEX \citep{galex-uvceti-flash}. The conclusion from DLS that M-dwarf flares constitute a dense ``foreground fog" in our Galaxy brings back a question on a derivation of the UV flare frequency rate, importance of which for other reasons was already discussed extensively in the literature \cite{....}.
In this article, we will discuss the possibility of \T in detecting the UV flares from M-dwarfs and from possible collisions in extra-solar planetary systems.
[edit] UV Emission from Celestial Collisions
The phenomena of one celestial body smashing into another is quite common on all astronomical scales, from asteroids bombarding planets to recently discovered collisions of massive galaxy clusters \cite{{bullet}, {cl0024}}. The origin of blue stragglers in the globular clusters is due to merging of two, or, may be three, main sequence (MS) stars of $\sim 0.8$ \msun. Perhaps about half the stars in central region of some globulars underwent one or two collisions over a period of $10^{10}$ years \citep{sivaram-ref3}. White dwarf (WD) binaries with periods $< 5$ mins will merge in a few thousand years, and neutron star--white dwarf (NS--WD) binaries with periods of $\sim 10$ mins have been observed \citep{sivaram-ref4}.
Over the past decades, a standard model for planetary system formation has emerged, in which the final stage is now strongly associated with a significant number of giant impacts on each of the young forming planets \cite{stern}. Giant Impact Models (GIM) are often invoked for explaining the formation of planetary systems, the Mars-sized terrestrial impactor believed to be responsible for the formation of the Moon was such an impact \cite{stevenson87}. However, even at later stages of planetary system life, the giant collisions are possible. In the past decade about 200 extra-solar planets (ESP) have been detected and they are a rapidly growing sample. Most of ESPs are giant hot jupiters, locked in a very short-period orbits with their star. It is reasonable to expect planetary and stellar-planetary collisions in such systems.
Collisions on all scales can be very energetic. Recent example in the Solar System is that of the comet Shoemaker-Levy that slammed into Jupiter. The fragment measuring 3 km across released $6\times 10^6$ megatons of energy, which is equivalent to one Hiroshima bomb every second continuously for ten years. If a WD smashes into a MS star like our Sun with an incoming velocity, say, of $\ge 700$ km/sec, the massive shock wave would compress and heat the Sun, the nuclear reactions will become much faster, and in about an hour, the Sun would release thermonuclear energy of about $10^{49}$ ergs, and the instabilities would blow the Sun apart in a few hours \citep{sivaram}.
In early days, when nothing was known about GRBs, they were thought to be of a galactic origin; one popular model being of an asteroid-size objects (with $\sim 10^{32}$ gm of mass) impacting a NS. In such case, with the impact velocity at a surface of a NS of $\sim 0.1 \,c$, the released kinetic energy would be $\sim 10^{52}$ ergs (or $T\sim 10^9$ K), with most radiation in the form of a few hundred Kev $\g$-rays. Recently, the impact theory was revived by proposing an alternative model for the short-duration GRBs, explaining them as supernovae followed the collisions of compact objects \citep{sivaram}.
[edit] Planet-planet collisions
As estimated by Stern (1994), for impacts with total collisional energy above $5\times 10^{34}$ ergs, the dominant heat sink will be radiation to space. We will consider here a Jovian planet (with mass $\sim M_{\rm J}$) being hit by a terrestrial-size planet (with mass $\sim m_{\rm E}$), as in such a case the total impact energy is of the order $E_{\rm imp} \sim GM_{\rm J}m_{\rm E}/R_{\rm J} \sim 10^{40}$ ergs. Giant impacts range from glancing events to direct head-on collisions. It is reasonable to assume hyper-velocity impacts, as, for example, in the end-member archetypes of planetary collisions. However, even the proposed origin of the Moon requires the impact velocity just barely above the two-body escape velocity \cite{nature-paper}. Still, even if a giant impactor approaches the target with negligible velocity, its decent into the target's \gv potential well will cause it to impact at, at least, near escape velocity. Therefore, in our calculations we assume the starting impact velocity as two-body mutual escape velocity, which for terrestrial and jovian-like bodies will be \be v^2_{\rm imp}\approx v_{\infty}^2+v_{\rm esc}^2\,, \label{eq:1} \end{equation} where $v_{\infty}$ is the random velocity of the encounting bodies. Assuming $v_{\infty}=0$, we obtain \be V_{\rm esc}=\sqrt{\fr{2G(M_{\rm J}+m_{\rm E})}{(R_{\rm J}+r_{\rm E}}}\approx 40 \, \text{km/sec}\,. \label{eq:2} \end{equation}
It was assumed in \citep{zhang} that because the energy deposit during the impactor descent is much greater than the corresponding Eddington luminosity of the target, $\dot E\approx E_{\rm imp}/\tau_{\rm shock} \approx 10^{38} \gg L^{\rm J}_{\rm Edd}= 10^{35}$erg/sec, the peak luminosity resulting from the hot spot created by the impactor will be $L_{\rm pk}\sim\eta L_{\rm Edd} \approx L^{\rm J}_{\rm Edd}$. Here $\tau_{\rm shock}$ is the shock crossing time for the impactor $\sim 600$ km/sec. This luminosity will be produced by the hot spot with the peak temperature of $\sim 10^5$ K, which corresponds to the electromagnetic flash at $\l \sim 450$ \AA. This is out of the TAUVEX detectors range (see Table~\ref{table:observatory}). However, it is not at all clear that $\eta$, the factor correcting for radiative inefficiencies in \citep{zhang}, shall be unity. Besides, the gas opacity of the atmosphere is crucial for the escape of radiation from the shocked matter, which was not considered in \cite{zhang}. Let us assume the atmospheric entry at velocity $v$ and angle $\phi$ (see Fig.~\ref{fig:entry} in App.~A). The incoming impactor generates a shock front in the atmopshere of a target with velocities in the range $10\,\text{km/s}\le v \le 40\,\text{km/s}$. The force of drag on an impactor is $F_{\rm D}=1/2 C_{\rm D} \rho_a A v^2$, where $\rho_a$ is atmospheric density and $C_{\rm D}$ is the coefficient of atmospheric drag ($C_{\rm D}\approx 1/7$ (for ex., \cite{chyba}). The power released due to the drag will be \be \dot E = \half C_{\rm D}\rho_a A v^3\,, \label{eq:3} \end{equation} where $A$ is the front area of the impactor. If we assume that all dissipated energy going into heat, $\eta\rho_a A v^3\sim \s A T_{\rm R}^4$, we obtain \be T_{\rm R} \sim \left(\fr{\eta\rho_a v^3}{\s}\right)^{1/4}\,, \label{eq:4} \end{equation} where $\s$ is the Stephan-Boltzmann constant and $\eta$ is a fraction of energy going into heat. We assume $\eta \approx 1$ ($T_{\rm R}$ it is relatively insensitive to it). The temperature estimates for different impact velocities are given in Table~\ref{tab:temperatures}. We see from the table that $T_{\rm R}$ strongly depends on the velocity of the impactor and wide range of temperatures is possible.
\begin{table*} \centering \caption{This table presents estimates of $T_{\rm R}$ in K from Eq.~\ref{eq:4} as a function of density $\rho$ (in g/cm$^3$) and impact velocity $v$.} \begin{tabular}{l|l|l} $\rho \setminus v$& 10 km/sec & 40 km/sec\\ \hline $10^{-8}$ & $\sim 3,600$ & $\sim 10,000$\\ $10^{-6}$ & $\sim 11,500$ & $\sim 32,500$\\ $10^{-4}$ & $\sim 36,000$ & $\sim 100,000$\\ $10^{-3}$ & $\sim 65,000$ & $\sim 180,000$\\ \hline \end{tabular} \label{tab:temperatures} \end{table*} For a thermal spectrum a region of $2-6\times 10^4$ K represents UV region. With $T_{\rm R}\sim 10^5$ the radiation is going into far-UV region, again inaccessible for \T\ns. However, the prompt radiation firstly strongly depends on initial velocity, and, secondly, as was detailed in \citep{sarazin}, as gas is optically thin in the top layers of atmospher, the initial radiation will be in emission lines. At $\approx 10^{-6}$, merely a $\sim$ second after start of descent, shocked atmosphere is optically thick, shock becomes a source of black-body continuum, and the favoured emmission is in $1000-3000$ \AA (UV range well covered by \T\ns) initially. This radiation can escape in the prompt phase from the vicinity of the shock front, but radiation bluer than $\sim 900$ \AA \; is trapped very close to the shock and all absorbed in a narrow preshock region (thus only longer UV radiation is capable to escape). At this prompt phase the luminosity can reach $L_{\rm UV} \sim 10^{31}$ erg/sec and the total integrated energy released during descent may reach $\sim 10^{32}$ ergs (see App.~A). Flux of UV photons from the Sun at the Earth (at 1 A.U.) is $\sim 3\times 10^{14}$ /cm$^2$/s, from a distance of 10 pc it will be $\sim 50$ phot/cm$^2$/s. For the UV flare from the collision described here the flux will be $\sim 2\times 10^3-5\times 10^3$ phot/cm$^2$/s for the duration from seconds \citep{sarazin} to 1-2 hours \cite{zhang}.
[edit] Stellar-planetary collisions
Let us consider here the system Sun-Jupiter. Many extra-solar jovian planets are in very tight orbits around the star. Out of total 246 (2007) discovered extra-solar planetray systems there are, at least, 30 planets with orbits less than 1 A.U. from the host. Assuming the velocities are of order $\sim 2000-300$ km/sec, the planet can spiral down in only $\sim 10^4$ years.
[edit] Estimates of frequency
. . . . . . .
[edit] M dwarf flares
bmjbiuubkb
M dwarfs account for more than 75\% of the stellar population in the solar neighbourhood (up to 1 Kpc). These stars are known to possess strong magnetic fields with high coronal activity and associated UV line emission [Mitra-Kraev et al 2005] and a vast majority of UV short transients is associated with a single physical type of astrophysical source, that is, stellar flare eruptions on K and M stars.
Importance of studies of dM flare frequency:
1. The level of chromospheric activity may vary with dM stellar age and mass [Gizis, Reid \& Hawley 2002]. There are observations [Welsh et al 2006 and refs within] that activity strength (as measured by the ratio of H$\a$ luminosity to the stellar bolometric luminosity) increases with the spectral class--- earlier dMs have higher flare energies. However, as was discussed in [Kulkarni \& Rau 2006], there may be a fraction of late-type dwarfs that retain their activity for a longer period or may have had activity induced later in their lives.
2. A derivation of the UV flare frequency rate on M dwarfs is also important for the study of habitability zones on possible associates ESP systems [Turnbull \& Tarter 2003].
3. Lastly, as has been esimated in [Kulkarni \& Rau 2006], the dM transients annual rate of $R_{\rm f} \sim 10&{8}\text{yr}^{-1}$ implies the existence of immense fog of M dwarfs that hinders any extragalactic transient search---the foreground fog of dM flares ensures that the false positives outnumber genuine extragalactic events by at least 2 orders of magnitude. The clear identification of all suspect dwarfs, especially over the extragalactic fields, will ensure that the extragalactic surveys, like, for ex. PANSTARRS or LSST, will not drown in false detections.
To write about:
1. Flare energy distribution peaks at $\l < 4000$ \AA [van den Oord et al. 1996], so optical photometry (usually used to observe flares) samples only a wing of this distribution (contaminated by the photosphere) and in most cases using only a single transmission filter. \T with $\l$ range ... and resolution of ... can simulteneously monitor $\sim 1^{\circ}$ field in three UV bands, defining two UV colours.
2. Principal filters in survey mode, favouring the serendipitious detection, are .... - Advantages of 2 UV colours.
3. Advantage of our pipeline, having time-tagged photon list for light curves.
GALEX pipeline only produces calibrated image of the field and thus
doesn't retain timing information.
4.
